GOLTC Data Science

• GOLTC Introduction Adelina Comas-Herrera
• Data Science Interest Group: Structure and webinars
  • Steering Group: Sam Rickman (CPEC at LSE), Jiyoun Song (UPenn), Tommy Henderson-Reay (NHS England), Sarthak Saluja (CPEC at LSE)
• Racial disparities and harmful hallucinations in automated speech recognition Allison Koenecke, Cornell Department of Information Science
• Adapted large language models can outperform medical experts in clinical text summarization [Dave Van Veen, Stanford Center for Artificial Intelligence]

Areas to cover

• Today: policy context of generative AI and its use in practice
• Future webinars: evaluating accuracy, bias and ethics
• General: resources and papers on GOLTC Data Science website

Generative AI in public services

UK Budget 2024

Public Sector Productivity Plan:

• £4.2 billion investment
• More efficient services
• Tech will free frontline workers from admin

By automating the writing and clinical coding of notes, discharge summaries and GP letters, clinicians will be able to spend more time with patients at more appointments

But how do we measure these models?

Today’s webinar: moderated discussion

• Practical questions
• Accuracy and bias
• Ethics and information governance
• Other areas
• Comments